Although the emergence of multi-parametric magnetic resonance imaging (mpMRI) has had a profound impact on the diagnosis of prostate cancers (PCa), analyzing these images remains still complex even for experts. This paper proposes a fully automatic system based on Deep Learning that performs localization, segmentation and Gleason grade group (GGG) estimation of PCa lesions from prostate mpMRIs. It uses 490 mpMRIs for training/validation and 75 for testing from two different datasets: ProstateX and Valencian Oncology Institute Foundation. In the test set, it achieves an excellent lesion-level AUC/sensitivity/specificity for the GGG$$\ge$$ ≥ 2 significance criterion of 0.96/1.00/0.79 for the ProstateX dataset, and 0.95/1.00/0.80 for the IVO dataset. At a patient level, the results are 0.87/1.00/0.375 in ProstateX, and 0.91/1.00/0.762 in IVO. Furthermore, on the online ProstateX grand challenge, the model obtained an AUC of 0.85 (0.87 when trained only on the ProstateX data, tying up with the original winner of the challenge). For expert comparison, IVO radiologist’s PI-RADS 4 sensitivity/specificity were 0.88/0.56 at a lesion level, and 0.85/0.58 at a patient level. The full code for the ProstateX-trained model is openly available at https://github.com/OscarPellicer/prostate_lesion_detection. We hope that this will represent a landmark for future research to use, compare and improve upon.
Prostate segmentations are required for an ever-increasing number of medical applications, such as image-based lesion detection, fusion-guided biopsy and focal therapies. However, obtaining accurate segmentations is laborious, requires expertise and, even then, the inter-observer variability remains high. In this paper, a robust, accurate and generalizable model for Magnetic Resonance (MR) and three-dimensional (3D) Ultrasound (US) prostate image segmentation is proposed. It uses a densenet-resnet-based Convolutional Neural Network (CNN) combined with techniques such as deep supervision, checkpoint ensembling and Neural Resolution Enhancement. The MR prostate segmentation model was trained with five challenging and heterogeneous MR prostate datasets (and two US datasets), with segmentations from many different experts with varying segmentation criteria. The model achieves a consistently strong performance in all datasets independently (mean Dice Similarity Coefficient -DSC- above 0.91 for all datasets except for one), outperforming the inter-expert variability significantly in MR (mean DSC of 0.9099 vs. 0.8794). When evaluated on the publicly available Promise12 challenge dataset, it attains a similar performance to the best entries. In summary, the model has the potential of having a significant impact on current prostate procedures, undercutting, and even eliminating, the need of manual segmentations through improvements in terms of robustness, generalizability and output resolution.
ImportanceFocal ablative irreversible electroporation (IRE) is a therapy that treats only the area of the tumor with the aim of achieving oncological control while reducing treatment-related functional detriment.ObjectiveTo evaluate the effect of focal vs extended IRE on early oncological control for patients with localized low- and intermediate-risk prostate cancer.Design, Setting, and ParticipantsIn this randomized clinical trial conducted at 5 centers in Europe, men with localized low- to intermediate-risk prostate cancer were randomized to receive either focal or extended IRE ablation. Data were collected at baseline and at regular intervals after the procedure from June 2015 to January 2020, and data were analyzed from September 2021 to July 2022.Main Outcomes and MeasuresOncological outcome as indicated by presence of clinically significant prostate cancer (International Society of Urological Pathology grade ≥2) on transperineal template-mapping prostate biopsy at 6 months after IRE. Descriptive measures of results from that biopsy included the number and location of positive cores.ResultsA total of 51 and 55 patients underwent focal and extended IRE, respectively. Median (IQR) age was 64 years (58-67) in the focal ablation group and 64 years (57-68) in the extended ablation group. Median (IQR) follow-up time was 30 months (24-48). Clinically significant prostate cancer was detected in 9 patients (18.8%) in the focal ablation group and 7 patients (13.2%) in the extended ablation group. There was no significant difference in presence of clinically significant prostate cancer between the 2 groups. In the focal ablation group, 17 patients (35.4%) had positive cores outside of the treated area, 3 patients (6.3%) had positive cores in the treated area, and 5 patients (10.4%) had positive cores both in and outside of the treated area. In the extended group, 10 patients (18.9%) had positive cores outside of the treated area, 9 patients (17.0%) had positive cores in the treated area, and 2 patients (3.8%) had positive cores both in and outside of the treated area. Clinically significant cancer was found in the treated area in 5 of 48 patients (10.4%) in the focal ablation group and 5 of 53 patients (9.4%) in the extended ablation group.Conclusions and RelevanceThis study found that focal and extended IRE ablation achieved similar oncological outcomes in men with localized low- or intermediate-risk prostate cancer. Because some patients with intermediate-risk prostate cancer are still candidates for active surveillance, focal therapy may be a promising option for those patients with a high risk of cancer progression.Trial RegistrationClinicalTrials.gov Identifier: NCT01835977
Aims The aim of this study was to compare magnetic resonance imaging (MRI) parameters in patients with mild incontinence after radical prostatectomy, who had undergone treatment with a suburethral sling. The objective was to compare patients who had been successfully treated with unsuccessful patients. Methods This observational cohort study at a single institution evaluated consecutive patients treated with an AdVance XP sling. MRI was performed using a 1.5 Tesla system. Preoperative urodynamic assessment and flexible cystoscopy were performed. The qualitative analysis was based on sling indentation (complete vs incomplete). The quantitative analysis comprised the following three parameters: the sling‐pubis distance, the thickness of the proximal urethral bulb, and the sling backward distance (SBD), defined as the distance between the prolongation of a line through the major axis of the pubis (the line‐segment joining the vertices of the pubis) and the sling indentation. The primary outcome was pad count at 3 months; cure as zero pads. A logistic univariate regression model was employed to assess the potential predictors of successful outcomes. An adjusted multivariate logistic regression model using the significant univariate factors was developed. Results Of the 83 patients enrolled, the univariate analysis revealed a relationship between failure and adverse urodynamics and between success and sling indentation, thickness of the proximal bulb and SBD. Only the association with SBD persisted in the multivariate analysis. Conclusions MRI revealed a strong relationship between proper positioning of the sling (small SBD) and continence outcome. These data have important implications for a second surgery following initial sling failure.
The emergence of multi-parametric magnetic resonance imaging (mpMRI) has had a profound impact on the diagnosis of prostate cancers (PCa), which is the most prevalent malignancy in males in the western world, enabling a better selection of patients for confirmation biopsy. However, analyzing these images is complex even for experts, hence opening an opportunity for computer-aided diagnosis systems to seize. This paper proposes a fully automatic system based on Deep Learning that takes a prostate mpMRI from a PCa-suspect patient and, by leveraging the Retina U-Net detection framework, locates PCa lesions, segments them, and predicts their most likely Gleason grade group (GGG). It uses 490 mpMRIs for training/validation, and 75 patients for testing from two different datasets: ProstateX and IVO (Valencia Oncology Institute Foundation). In the test set, it achieves an excellent lesion-level AUC/sensitivity/specificity for the GGG≥2 significance criterion of 0.96/1.00/0.79 for the ProstateX dataset, and 0.95/1.00/0.80 for the IVO dataset. Evaluated at a patient level, the results are 0.87/1.00/0.375 in ProstateX, and 0.91/1.00/0.762 in IVO. Furthermore, on the online ProstateX grand challenge, the model obtained an AUC of 0.85 (0.87 when trained only on the ProstateX data, tying up with the original winner of the challenge). For expert comparison, IVO radiologist's PI-RADS 4 sensitivity/specificity were 0.88/0.56 at a lesion level, and 0.85/0.58 at a patient level. Additional subsystems for automatic prostate zonal segmentation and mpMRI non-rigid sequence registration were also employed to produce the final fully automated system. The code for the ProstateX-trained system has been made openly available at https://github.com/OscarPellicer/ prostate_lesion_detection. We hope that this will represent a landmark for future research to use, compare and improve upon.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.